Slug Generator

What each character becomes

The field is normalized with NFKD, then every combining mark in the range U+0300 through U+036F is deleted, then the string is lowercased. café loses the acute and becomes cafe. naïve becomes naive. Letters that do not split into a base letter plus a mark are not rescued. ß stays ß through that step and is then deleted, so straße becomes strae. æ is deleted, so Æther becomes ther. The page does not have a transliteration table for those letters.

After lowercasing, every character outside a-z, 0-9, whitespace, and hyphen is removed. An underscore is outside that set, so it vanishes: Hello_World becomes helloworld, and foo_bar_baz becomes foobarbaz if there are no spaces. A hyphen that you typed is kept, then runs of whitespace, underscores that somehow remained, and hyphens are collapsed to one hyphen. In practice the underscore is already gone, so the collapse is about spaces, tabs, and hyphens. The ends are trimmed of hyphens. ---Hello, World!--- becomes hello-world. 100% becomes 100. An emoji is deleted, so emoji 🎉 title becomes emoji-title.

The live field

There is one textarea and one result labeled Slug. Every change recomputes. You cannot pick an underscore separator, a maximum length, or a stop-word list. The words for, to, and a stay. 10 SEO Tips for Better Metadata keeps for and better. Case Converter can show a kebab line too, but it keeps commas and it turns underscores into hyphens. Those are different rules. Paste the title here when the destination is a URL segment.

An empty field publishes an empty string, so Copy output says there is nothing to copy, while the panel still shows No slug yet. A field that collapses to nothing, such as a line of only punctuation, behaves the same way. A non-empty slug is the copy buffer, with no added newline and no site hostname in front of it. The page does not build a full URL.

Jobs for a URL piece

Use the slug as the path piece, then build the rest of the address elsewhere. Campaign parameters belong on the UTM builder, which will not slugify the path for you. A title tag and a description for that URL belong on the meta tag generator. If the path must be percent-encoded because a consumer still wants the raw characters, that is URL encode / decode, and it is a different result from a hyphenated slug.

The case converter is the right neighbor when you also need camelCase or CONSTANT_CASE from the same title. Its kebab panel is not this slug. Don't stop becomes dont-stop here, because the apostrophe is deleted and the space becomes a hyphen. On the case converter the apostrophe would remain inside the word.

Tabs, accents, and letters that disappear

A tab is whitespace, so a and b separated by a tab become a-b. Several spaces do the same. A hyphen next to those spaces collapses with them: foo - bar becomes foo-bar. Leading and trailing hyphens produced by the cleanup are removed, so a title that is only hyphens ends as the empty slug and the panel shows No slug yet.

İstanbul becomes istanbul because the dotted capital I splits under NFKD and the extra mark is removed before lowercasing. That is not a promise for every non-Latin script. Characters that are not a-z after this cleanup are dropped rather than approximated. If the whole title is in a script that does not decompose into a-z, the slug is empty. The page will not fall back to a percent-encoded path.

Rules and limitations

NFKD, strip combining marks U+0300–U+036F, lowercase, delete characters other than a-z, digits, whitespace, and hyphen, collapse whitespace and hyphens, then trim hyphens. Underscores are deleted.

The slug updates on each change. There is no separator choice, no stop-word list, and no length cap. Copy output is the slug, or nothing when the slug is empty. The panel text No slug yet. is not copied.

Common problems

If an underscore vanished, that is the rule. foo_bar becomes foobar. Put a space or a hyphen in the title if you wanted a break.

If ß or æ disappeared, the character did not decompose into a-z. café works because é does decompose. This page will not map ß to ss.

Frequently asked questions

About this tool

Maintained by Toolbench. For large or deeply nested files, split the input into smaller batches and verify representative results before using them in an import or application. Report reproducible problems through Contact.